Brain reactive antibodies (BRAs) have been identified in the blood of a majority of patients with Alzheimer's disease (AD). However, the role of these antibodies in the pathogenesis of the disease is unknown. Preliminary research from our laboratory suggests that BRAs occurs more frequently in AD patients than normal controls and that these antibodies have a different pattern of activity between patients with Alzheimer's disease and that of adult patients with Down's syndrome. These data may also have implication for the development of possible subtypes of AD. This application proposes to test a number of hypotheses for establishing the specificity, activity, and function(s) of BRAs in Alzheimer's disease. Sera from 240 patients with Alzheimer's disease, 90 patients with multi-infarct dementia, (age 60 or older), 90 elderly normal subjects, and 30 patients with Down syndrome, 35 years or older, will be analyzed for the occurrence, quantification, and consequently, that neuronal cell loss in Alzheimer's disease may be mediated by antibodies. Samples of brain obtained at autopsy of Alzheimer's cases will be eluded for determining the occurrence of brain-bound antibodies. Electroblots prepared from spinal cord, samples of brain and specific proteins isolated from brain will be used for testing the antigenic specificity of BRAs and the value of BRAs as markers in identifying possible subtypes of AD. Data will also be generated to examine the hypotheses a) that the occurrence of BRAs is related to the duration, age of onset, and family history of dementia; and b) that the pattern of cognitive loss determined using a battery of information processings is related to the presence of circulating BRAs.